Myocardial ischemia caused by obstruction of coronary arteries is often treated with percutaneous coronary intervention (PCI) where the obstruction is dilated and one or more stents are inserted to maintain the passage of blood. In addition to usually life-long treatment with aspirin, dual antiplatelet therapy with a P2Y12-antagonist is recommended for 12 months after PCI to counteract thromboembolic events, including occlusion of stents which could otherwise cause renewed ischemia and infarction[1, 2].

Proton pump inhibitors (PPI), one of the most common groups of medications globally, reduce the acidity of the gastric contents and are mainly used in the treatment or prevention of gastro-esophageal reflux disease or peptic ulcer disease. PPI is also recommended for gastric mucosa protection in patients with an elevated risk of duodeno-gastric bleeding who require antiplatelet therapy. Some studies have suggested that PPI treatment reduces the efficacy of clopidogrel [3], which may be due to shared cytochrome P450 enzymes (including CYP2C19) and reduced metabolism of the prodrug clopidogrel to the active metabolite (2-oxo-clopidogrel) of clopidogrel [4]. Such reduction in the efficacy of clopidogrel could potentially increase the risk of thromboembolic events after PCI. Existing data on the topic are conflicting, however, which highlights the need for large and population-based studies [3, 5].

The aim of this study was to help clarify if and to what extent concomitant use of PPI increases the risk of major thromboembolic events among patients who use clopidogrel after PCI in a large and unselected cohort.

This study found increased risks of myocardial infarction, coronary heart disease, stroke, and all-cause and disease-specific mortality among patients receiving PPI treatment together with clopidogrel compared to those only using clopidogrel within 12 months of PCI.

Among methodological strengths of the study are the population-based and nationwide design and the complete follow-up of participants, which provided a large sample size, minimized selection bias and losses to follow-up, and facilitated generalizability. Another advantage was the adjustment for several confounders, including obesity-related diagnoses and tobacco-related diagnoses. Yet, a limitation was that residual confounding by these indirectly measured exposures cannot be excluded, and these factors might still contribute to the associations. Additionally, the influence of some other potentially relevant confounders was not possible to adjust for, e.g., lifestyle factors, dietary exposures, or physical activity; we cannot rule out that PPI is only a proxy for other unmeasured factors leading to a reduced risk of cardiovascular events. Another limitation was the lack of data on adherence to medication with clopidogrel and PPI or whether the medication was discontinued on medical grounds. Adherence to clopidogrel should be high given the serious risks associated with non-use after PCI, while the continuing need for PPI use is less certain because not all patients will receive it as primary gastric protection but may use it for treatment of gastroesophageal reflux disease. PPI is available over-the-counter but only in small packages and at a much higher cost for the patient. Thus, patients requiring long-term treatment with PPI receive this through prescriptions. Any discontinuation of PPI use or missed PPI treatment due to over-the-counter medication should occur at random, and thus not explain the associations found, but rather dilute them.

Whether the use of a PPI together with clopidogrel increases the risk of cardiovascular events after PCI compared to the use of clopidogrel alone is a matter of debate. A recent systematic review and meta-analysis of 23 studies and 156,823 patients found that concomitant use of PPI and clopidogrel increased the risk of myocardial infarction (risk ratio 1.43, 95% CI 1.24–1.66), other major adverse cardiovascular events combined (risk ratio 1.22, 95% CI 1.06–1.39), and cardiovascular death (risk ratio 1.21, 95% CI 0.97–1.50) [3]. However, a sub-analysis of six randomized controlled trials (4476 patients) in that meta-analysis showed no increased risk of major cardiovascular events (risk ratio 0.99, 95% CI 0.76–1.28) [3]. The previously largest original study on the topic, a cohort study of 24,702 patients with myocardial infarction, showed that compared to non-users of PPI, the risk of major cardiovascular events was similarly increased among concomitant users of PPI and clopidogrel (HR 1.29 (95% CI 1.17–1.42)) as in PPI users without clopidogrel (HR 1.29 (95% CI 1.21–1.37)) but that study did not only include patients undergoing PCI [5]. The largest randomized controlled trial on the topic included 3761 patients with any indication for clopidogrel and found no difference between the group receiving clopidogrel and PPI and the group receiving only clopidogrel (HR 0.99 (95% CI 0.68–1.44)) [14], but the statistical power was low and the duration of follow-up was short (median of 106 days) [14]. Another observational study including 706 patients, among which approximately two-thirds received clopidogrel, found that concomitant use of PPI was associated with an increased risk of acute ischemic events, heart failure, or death (HR 2.28 (95% CI 1.24–4.18)) [15]. The present study, focusing on cardiovascular events specifically after PCI, is substantially larger than any previous original study on the topic and adds evidence in favor of an increased risk of major cardiovascular events if patients use PPI together with clopidogrel after PCI.

The proposed mechanism behind the associations found in this study is decreased effectiveness of clopidogrel when PPI is administered. Several in vitro studies have shown that PPI and clopidogrel share the metabolic pathway through the cytochrome P450 enzyme (CYP2C19), an enzyme necessary for the metabolism of clopidogrel to its active metabolite [16]. An in vitro study suggested that different types of PPI may affect specific CYP2C19s differently, with pantoprazole being the least potent inhibitor and lansoprazole the most potent inhibitor [17]. This finding prompted our separate analyses of the three most common types of PPI, but these analyses showed no obvious differences in risk estimates. Additionally, a portion of the population is carriers of an allele leading to reduced function of CYP2C19, which has been estimated to affect approximately 30% of the adult Caucasian and Black populations and 60% of Asian populations, and can lead to reduced effectiveness of clopidogrel [18, 19]. Intermediate or poor metabolizers are expected to be represented in the current cohort as well and might have affected the results.

Both the American College of Cardiology expert consensus and the European Society of Cardiology recommend the use of PPI concomitant to treatment with clopidogrel among patients with a history of upper gastrointestinal bleeding as well as among patients with multiple risk factors for gastrointestinal bleeding [2, 20]. Except for these indications of using PPI, the results of the present study suggest that PPI should be used cautiously among patients treated with clopidogrel after PCI, particularly in patients with less clear indications for PPI treatment.

In conclusion, the findings of this large and population-based cohort study indicate that the large group of patients who use PPI together with clopidogrel following PCI are at an increased risk of serious cardiovascular events compared to patients who use clopidogrel only after this intervention. The findings argue in favor of avoiding less necessary PPI treatment during clopidogrel treatment within 12 months of PCI.